Artificial Intelligence and Postoperative Acute Kidney Injury

Development and Prospective Validation of an Artificial Intelligence Model to Predict Postoperative Acute Kidney Injury

The main objective of this study is to develop and validate an artificial intelligence model that predicts postoperative acute kidney injury.

Postoperative acute kidney injury is known to increase the length of hospital stay and healthcare cost. A lot of risk prediction models have been developed for identifying patients at increased risk of postoperative acute kidney injury. Recent advances in artificial intelligence make it possible to manage and analyze big data. Prediction model using an artificial intelligence and large-scale data can improve the accuracy of prediction performance. Furthermore, the use of an artificial intelligence may be a useful adjuvant tool in making clinical decisions or real-time prediction if it is integrated into the electrical medical record systems. However, before implementing an artificial intelligence model into the clinical setting, prospective evaluation of an artificial intelligence model's real performance is essential. However, to our knowledge, there was no artificial intelligence model for prediction of postoperative acute kidney injury, which was prospectively evaluated. Therefore, we aimed to develop an artificial intelligence model which predicts postoperative acute kidney injury and evaluate the model's performance prospectively.

Inclusion Criteria:

• Adults patients undergoing non-cardiac surgery

Exclusion Criteria:

• Age under 18 years
• Surgery duration < 1 hour
• Transplantation surgery
• Nephrectomy
• Cardiac surgery
• Patients who had severe kidney dysfunction preoperatively as follows:
• Serum creatinine ≥ 4 mg/dl
• Estimated glomerular filtration rate <15 ml/min/1.73m2
• History of renal replacement therapy
• Patients who had no results of preoperative or postoperative serum creatinine